CN115294080B - Automatic slotting robot for highway cracks and working method and application thereof - Google Patents

Automatic slotting robot for highway cracks and working method and application thereof Download PDF

Info

Publication number
CN115294080B
CN115294080B CN202210974154.5A CN202210974154A CN115294080B CN 115294080 B CN115294080 B CN 115294080B CN 202210974154 A CN202210974154 A CN 202210974154A CN 115294080 B CN115294080 B CN 115294080B
Authority
CN
China
Prior art keywords
image
crack
slotting
robot
center
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202210974154.5A
Other languages
Chinese (zh)
Other versions
CN115294080A (en
Inventor
孔奥星
李济正
高飞
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shandong University
Original Assignee
Shandong University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shandong University filed Critical Shandong University
Priority to CN202210974154.5A priority Critical patent/CN115294080B/en
Publication of CN115294080A publication Critical patent/CN115294080A/en
Application granted granted Critical
Publication of CN115294080B publication Critical patent/CN115294080B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/70—Denoising; Smoothing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/136—Segmentation; Edge detection involving thresholding
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10004—Still image; Photographic image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20021—Dividing image into blocks, subimages or windows

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Quality & Reliability (AREA)
  • Image Processing (AREA)

Abstract

The application relates to an automatic slotting robot for highway cracks and a working method and application thereof, belonging to the technical field of highway maintenance equipment, and comprising a robot trolley, wherein a navigation camera, a head light source and a laser radar are arranged at the front part of the robot trolley, a crack image acquired by the navigation camera is extracted from the center and then used for the navigation of a walking route of the robot, a slotting cutter, a bottom camera and a bottom light source are arranged below a chassis of the robot trolley, and the slotting cutter rotates at a high speed to complete the slotting of a highway during working; the slotting cutter is driven by the slotting cutter driving mechanism, and the bottom camera is used for shooting a crack image of the bottom area of the trolley in a trolley parking state, and guiding slotting props to slotting after the center line track is extracted from the crack image. The application can realize automatic tracking and slotting of highway cracks, has high automation degree and intelligent degree and slotting precision, and has very high expansibility and compatibility for integrated highway repair equipment.

Description

Automatic slotting robot for highway cracks and working method and application thereof
Technical Field
The application relates to an automatic slotting robot for highway cracks, a working method and application thereof, and belongs to the technical field of highway maintenance equipment.
Background
Road cracks are one of the common road hazards, the cracks can be gradually lengthened and widened along with rolling of vehicles, comfort and safety in the running process of the vehicles can be affected, accumulated water can infiltrate into roadbeds through the cracks, the service life of the road is greatly reduced, and even sedimentation or collapse of the road can be caused. Along with the great number of highways in China entering into maintenance period, the repair of the highways cracks becomes a great difficulty in highway maintenance.
Road cracks are of various forms, in which the ratio of longitudinal and transverse cracks parallel and perpendicular to the direction of road extension is the highest, and the effect is the most serious, so automatic repair techniques for studying these cracks have been left alone. At present, the repairing method for the road cracks is divided into a traditional method and a grooving repairing method. The traditional method is to directly pour the repairing filler into the crack, and has the advantages of convenient operation and higher efficiency; the defect is that the filler cannot be fully filled when sundries exist in the cracks or the width of the cracks is too small, so that the repair is not thorough and the effect is poor. The grooving repairing method is to process a bottom groove with certain width and depth along the center line of the crack by a grooving cutter, and then to fill repairing filler into the groove. Compared with the traditional method, the repairing filler can be fully contacted with the crack (bottom groove), so that the crack is fully repaired, and the repairing effect is good. In the two steps of slotting and filling, slotting of a crack is the most important step, and slotting quality determines the effect of crack repair.
The main ways of realizing highway crack slotting are manual slotting, hand-held equipment slotting, simple slotting machine slotting and the like, wherein the simple slotting machine is most widely applied in large-scale application. The grooving machine is generally provided with grooving cutters (grooving saw, grooving cutter and the like), wherein the cutters are fixed on a lathe and driven by a fuel oil engine or a motor to rotate at high speed; during grooving operation, the grooving machine is manually pushed or pulled (or auxiliary power) to walk along the center track of the crack through visual observation, and the advancing direction is continuously adjusted so as to enable the rotary cutter to pass through the center of the crack to finish grooving. Such slotting devices, although of relatively simple construction, have a number of problems, mainly: 1) The labor intensity of workers is high, the slotting efficiency is low, and the slotting process has larger dust and noise pollution, so that the health of the workers is seriously influenced; 2) The slotting precision is low and the quality is unstable. The grooving quality is influenced by the operation proficiency and mental state of workers, and the problems of large deviation between the grooving center and the crack center and the like often occur; in addition, the grooving quality is completely dependent on an individual operator, so that the quality is unstable and the precision consistency is poor; 3) The automation and the intellectualization of the equipment are insufficient, and the modern process of highway maintenance is seriously affected. Currently, there are few mature solutions to the problem of automatic slotting or repair of highway cracks.
Disclosure of Invention
Aiming at the defects of the prior art, the application provides an automatic slotting robot for highway cracks, a working method and application thereof.
The technical scheme of the application is as follows:
the automatic slotting robot for the highway cracks comprises an Ackerman robot trolley, wherein a chassis of the robot trolley comprises two rear wheels driven by differential speed and two front wheels for controlling steering;
the front part of the robot trolley is provided with a navigation camera, a head light source and a laser radar, a crack image acquired by the navigation camera is extracted from the center and then used for navigation of the robot, the head light source is used for navigation illumination, and the laser radar is used for automatic obstacle avoidance of the robot trolley;
a slotting cutter, a bottom camera and a bottom light source are arranged below the chassis of the robot trolley, the slotting cutter is a tooth saw or a milling cutter, and the slotting cutter rotates at a high speed to complete slotting of a highway during working; the grooving cutter is driven by the grooving cutter driving mechanism and can be driven to perform linear motion along the X/Y/Z directions while the grooving cutter is driven to rotate at a high speed; the X/Y linkage is used for processing and obtaining a continuous track of a crack center, the Z direction is used for controlling the depth of a groove, the bottom camera is used for shooting a crack image of a vehicle bottom area in a vehicle parking state, and the bottom light source provides illumination for the bottom camera.
To realize automatic navigation of the slotting robot trolley, two working modes are respectively designed, wherein one working mode is a continuous slotting working mode, as shown in fig. 3-1; the other is the intermittent slotting mode, as shown in fig. 3-2.
(1) The continuous mode of operation is the most intuitive and common mode of operation. The mode simulates manual slotting equipment, and the slotting robot trolley performs slotting operation while performing line inspection navigation; line patrol navigation is realized by a head camera; the running track of the slotting cutter is the center track of the crack. The slotting cutter does not require a cutter position adjustment mechanism in addition to the rotary primary motion.
The working mode has the advantages that the robot trolley is simple in structure and high in slotting efficiency; the disadvantage is that it is difficult to meet the grooving accuracy requirement for the following reasons:
the main factors influencing slotting accuracy are line inspection errors, which are mainly influenced by two parts, namely an imaging and center extraction algorithm error delta 1 of a navigation camera, which is generally smaller; secondly, the walking error delta 2 of the robot trolley is affected by factors such as the weight, inertia, ground friction, vibration and the like of the slotting trolley, the walking error delta 2 is far greater than delta 1, and in general, the error delta 2 only can far exceed the slotting precision requirement. And the actual slotting accuracy error is the sum of the two parts. Therefore, the working mode is difficult to meet the grooving precision requirement.
(2) The intermittent slotting working mode of the application is shown in fig. 3-2. The robot trolley is navigated by the head navigation camera, after the robot trolley advances by one vehicle body length, the trolley is stopped, the bottom camera shoots, a crack image covered by the bottom of the trolley is obtained, the center of a track is extracted, and the center track provides a feeding instruction in the X/Y/Z direction for a slotting cutter driving mechanism, so that slotting is completed along the center of the crack by the slotting cutter.
The working mode has the defects that the structure of the robot trolley is relatively complex, and the slotting efficiency is relatively low; the grooving device has the advantages of high grooving precision, and the reason is as follows:
under the working mode, the grooving operation is performed after the line inspection is performed, and the trolley is stopped, so that the cruise camera error delta 1 and the robot walking error delta 2 do not influence the grooving precision. The slotting precision is affected by two parts, namely, the error delta 3 of bottom camera imaging and crack center detection is small; secondly, under the existing technical conditions, even a low-cost three-dimensional movement mechanism can achieve high movement precision, so that the error delta 4 is small. Therefore, the intermittent working mode designed by the application has the grooving precision far higher than that of the continuous working mode, and can meet the grooving requirement of highway cracks, so that the intermittent working mode is adopted by the application.
Aiming at the grooving working mode and the robot structural design, the working method of the automatic grooving robot for the highway cracks comprises the following steps:
(1) Guiding the slotting robot trolley to a crack starting position, so that a crack is positioned between wheels on two sides and is positioned in a field of view of a navigation camera;
(2) The robot trolley starts a slotting mode;
(3) A navigation camera at the front part of the trolley starts to collect a road surface image at the front part of the trolley, extracts a center line of a crack and obtains a center track image of the crack;
(4) The robot automatically detects the deviation between the current pose and the center position of the crack, automatically tracks the center of the crack, and stops after controlling the trolley to move forward by one vehicle body distance;
(5) Starting a bottom camera by the robot trolley, shooting a bottom crack image, extracting a crack center line, and obtaining a crack center track image of a vehicle bottom area;
(6) The robot controls the slotting cutter driving mechanism to drive the slotting cutter to move, slotting is completed along the center track of the crack obtained in the step (5), and the slotting cutter position is reset (the initial position in the X/Y/Z three-dimensional direction); the driving mechanism can avoid the defects of poor rigidity, large slotting vibration, complex control, high cost and the like of the mechanical arm.
(7) Judging whether a crack center line exists in the visual field of the navigation camera, if yes, returning to the step (3), continuing the slotting work of the next parking space, and realizing autonomous navigation through a crack center line extraction method and a crack center automatic tracking method; if the navigation camera has no crack center line in the field of view, the robot stops working and sends task completion alarm information. The present application utilizes a front one of the cameras dedicated to navigation and a bottom one of the cameras dedicated to tool path generation. The interference is avoided, the method is simple and the precision is high.
Preferably, in the step (3), the crack center line extraction method includes the steps of:
(3-1) reading in the road crack image acquired by the navigation camera, and converting the road crack image into a gray image, specifically:
f(u,v)＝0.2989*r(u,v)+0.587*g(u,v)+0.114*b(u,v) (1)
wherein f (u, v) is a gray scale image matrix; r (u, v), G (u, v), B (u, v) are R/G/B component matrices of RGB images, respectively;
(3-2) gaussian filtering the slit gray image:
F(u,v)＝G(u,v)*f(u,v) (2)
wherein ,the method is characterized in that the method is a Gaussian filter kernel function, u is the number of lines of an image, v is the number of columns of the image, and sigma is the standard deviation of the kernel function; the filter size is 7x7, the standard deviation is sigma=0.5-3.5, and the standard deviation is preferably sigma=2;
(3-3) performing threshold segmentation on the filtered image to obtain a binary image, wherein the specific method comprises the following steps of:
wherein t=255×0.1=25.5;
(3-4) carrying out morphological processing on the binary image, removing noise points, and obtaining a crack center image, wherein the specific process is as follows:
1) The 8 connected domain marking is carried out on the binary image h (u, v), specifically:
let (u, v) be the center pixel, search and mark the connected domain for 8 neighborhood including (u+1, v), (u-1, v), (u, v+1), (u, v-1), (u+1, v+1), (u-1, v+1) and (u-1, v-1), the specific method is:
a. progressive scanning image, comprising a sequence of successive white pixels in each row, each sequence being assigned a reference j i ，i＝1,2，3...；
b. Judging whether the sequence of the line is communicated with the sequence in the previous line one by one from the second line of the image, namely judging whether each white pixel point in the sequence has a white point in the 8 neighborhood of the previous line, if not, representing non-communication, and if so, representing communication; if not, it is given a new reference (the largest reference in the previous row is j i Then the new reference sign is j i+1 ) The method comprises the steps of carrying out a first treatment on the surface of the If it is connected to only 1 sequence in the previous row, the sequence is combined with the sequence of the previous row to form a new sequence, and the new sequence is given the reference number of the sequence of the previous row (the previous row has two sequences in total and the reference numbers are j respectively i and ji+1 If a certain sequence of this row is numbered j in the previous row i Is to connect two rows in a middle connectionThe concatenated sequences are combined and labeled j i Assign a new sequence); if the sequence is communicated with 2 or more sequences in the previous row, combining the sequence with all the communicated sequences to form a new sequence, and assigning the smallest number in the sequence in the previous row to the new sequence (the two sequences in the previous row are respectively numbered j i and ji+1 If a sequence in the row is connected with both sequences in the previous row, all connected sequences in the two rows are combined, and the smallest number, i.e. j i Assigned to a new sequence) and writing the labels of the sequences already connected in the previous row to the equivalent pair < j i ,j i+1 ,. >, indicating that the sequences marked by these numbers belong to the same sequence;
c. and judging all equivalent pairs, and if the two equivalent pairs have the same reference numerals, merging the two equivalent pairs into a new equivalent pair because the sequences marked by the two equivalent pairs belong to the same connected domain. Finally, each new equivalent pair is marked with a connected domain label, denoted as L i The method comprises the steps of carrying out a first treatment on the surface of the At the same time, the number of connected domain pixels marked by the label of each connected domain is recorded and is recorded as N i The method comprises the steps of carrying out a first treatment on the surface of the Wherein i=1..m is the number of connected domains in the image; then n= [ N ] 1 N 2 ...N m ]Representing the number of pixel points of each connected domain in the image; with L= [ L ] 1 L 2 ...L m ]A reference numeral representing a corresponding one of the connected domains;
2) According to the number N of the pixel points of each connected domain i The communicating domains are ordered in descending order, [ N ', L ]']＝Rank(N,'descend')
Wherein, the number N of each connected domain pixel in N' is i Descending order from big to small; the element in L' is each N i The labels of the corresponding connected domains; 'desend' represents a descending order;
3) The first 10 connected domains N with the largest pixel number are reserved 1 ～N 10 I.e. in N 10 For the segmentation threshold, when the number of the pixels of the connected domain is larger than the segmentation threshold, reserving the connected domain; if the number of the pixels of the communication area is smaller than the segmentation threshold value, discarding the communication area; after the self-adaptive threshold segmentation step, obtaining cracksA binary image H (u, v);
4) Performing closed operation to connect broken cracks; the specific operation steps are as follows:
the image H (u, v) is inflated by a disk structure S with a radius of 5 pixels, and then etched,
wherein ,Θ is the image expansion and corrosion symbol, respectively;
5) Performing skeleton extraction on the image obtained in the step 4), and extracting a central pixel point of a crack, wherein the method comprises the following specific steps of:
assuming that the image matrix obtained through the above closed operation is H (u, v), the crack center image can be obtained by the following method:
wherein ,is a subset of the fracture center image, which can be obtained by combining the subset. Wherein Θ, & gt>Respectively image erosion and open operation symbols; (H Θkb) represents that the image H (u, v) is subjected to successive k times of etching with the structure B, which is a circular structure having a radius of 7 pixels; then, performing open operation on the corroded image, eliminating a small white area, and solving the difference between the corroded image and the corroded image, so as to obtain a crack center subset image; preferably, the number of times of k is determined by: is the last iteration before H (u, v) is eroded to empty set, i.eThen, the image S (H) processed as described above represents a skeleton image of the image H (u, v), denoted as S (u, v), representing an image matrix of the center of the crack,
preferably, in the step (4), the automatic tracking method for the center of the crack comprises the following steps:
setting O W -X W Y W Z W Is the world coordinate system, O V -X V Y V Z V Is the car coordinate system, Z W and ZV The axes being respectively with O W -X W Y W and OV -X V Y V The plane is vertical; o-uv is the image coordinate system; the image S (u, v) relates to X V Axisymmetric, and the image closest to the row and X of the trolley V The intersection point of the axes is denoted as X;
(4-1) extracting a center line image S (u, v) by using the crack center line extraction method in the step (3), wherein any point on the image is defined as S (u, v);
(4-2) at any time, taking the intersection point S of the crack center and the image center line 1 (u 1 ,v 1 ) The method comprises the steps that as a current navigation tracking point, a heading angle corresponding to the point is phi, the point is defined as a tracking starting point, and an image center line is parallel to a v axis along a u axis;
(4-3) to achieve crack center tracking, it is required to control the speed V of the cart X Along X V By controlling the angular velocity omega of the heading while the shaft is advancing Z The tracking start point is made to fall at the center S 'of the image S (u, v)' 1 (u 1 ,v 1 ) The point is defined as a navigation tracking target point;
(4-4) in actual control, since the forward speed V of the carriage X Is known and adjustable about Z V Heading angular velocity ω of shaft Z Is determined by the following method:
let it be assumed that the start point S is tracked 1 (u 1 ,v 1 ) There is a mapping point S 0 (u 0 ,v 0 ) Wherein |XS 0 |＝|XS 1 I, the distance from the mapping point to the tracking start point to the X point is the same; assuming that the tracking process is completed within Δt time, thenMeaning that the trolley is along X for a time Δt V The axis is moved by a distance |S '' 1 S 0 Simultaneous with I along Z V The heading angle is phi by rotating the shaft, and the heading angular velocity can be calculated by the following formula:
wherein, |XS' 1 The actual length of the I in the world coordinate system can be obtained through calibration of an imaging system;
(4-5) controlling start and stop of the robot trolley according to the following method: as long as the crack centerline trajectory can be detected in the crack centerline image, the robot trolley always determines V according to the methods of steps (4-1) - (4-4) X and ωZ Advancing; when no crack central line track exists in the visual field image shot by the trolley navigation camera, V is determined X and ωZ All set to 0 and the robot trolley is stopped.
The utility model provides an automatic slotting robot of highway crack is in application of integration highway repair system, integration highway repair system includes automatic slotting robot of highway crack, crack filling device, and automatic slotting robot of highway crack is used for carrying out automatic slotting to the highway crack, and crack filling device is used for the crack filling after the slotting.
The application has the beneficial effects that:
(1) The slotting precision is high. Under the condition of low-cost configuration of a robot trolley, a camera, a slotting movement mechanism and the like, the slotting center positioning precision can be controlled to be more than +/-5 mm.
(2) The automation degree and the intelligent degree are high. The automatic tracking and slotting of the highway cracks can be realized, the automatic walking and automatic slotting mode is adopted, the universality on transverse, longitudinal and oblique cracks is high, the slotting efficiency is improved, the labor intensity of workers is reduced, and the highway maintenance cost is reduced. The hardware and software algorithms for navigation and slotting track generation are separated, complex robot map building and coordinate transformation are not involved, and the control structure and algorithm are simple.
(3) And the expandability is strong. Based on the automatic slotting robot implementation scheme and the visual navigation method, the automatic slotting robot is provided with a crack pouring device and other auxiliary devices, so that integrated highway crack repairing equipment integrating slotting and crack pouring is easy to construct.
Drawings
FIG. 1 is a schematic perspective view of a highway crack automatic grooving robot;
FIG. 2 is a schematic diagram of the bottom structure of the automatic slotting robot for highway cracks according to the present application;
FIG. 3-1 is a schematic diagram of a continuous slotting operation of a robot;
FIG. 3-2 is a schematic diagram of a robot intermittent slotting operation;
FIG. 4 is a schematic diagram of a grayscale image of a road crack image after grayscale processing;
FIG. 5 is a schematic representation of an image after Gaussian filtering of a gray scale image;
FIG. 6 is a schematic diagram of a binary image after processing a filtered image;
FIG. 7 is a schematic view of a binary image of a crack after an adaptive threshold segmentation step;
FIG. 8 is a schematic image of a closed-loop operation;
FIG. 9 is a schematic view of an obtained crack center image;
FIG. 10 is a schematic diagram of a coordinate system setup of a crack center tracking process;
wherein: 1. the laser radar system comprises a head light source 2, a laser radar 3, a navigation camera 4, a slotting prop 5, a slotting cutter driving mechanism 6, a robot trolley chassis 7, a bottom camera 8 and a bottom light source.
Detailed Description
The application will now be further illustrated by way of example, but not by way of limitation, with reference to the accompanying drawings.
Example 1:
the automatic slotting robot for the highway cracks comprises an Ackerman robot trolley, wherein a chassis of the robot trolley comprises two rear wheels driven by differential speed and two front wheels for controlling steering;
the front part of the robot trolley is provided with a navigation camera, a head light source and a laser radar, a crack image acquired by the navigation camera is extracted from the center and then used for navigation of the robot, the head light source is used for navigation illumination, and the laser radar is used for automatic obstacle avoidance of the robot trolley;
a slotting cutter, a bottom camera and a bottom light source are arranged below the chassis of the robot trolley, the slotting cutter is a tooth saw or a milling cutter, and the slotting cutter rotates at a high speed to complete slotting of a highway during working; the grooving cutter is driven by the grooving cutter driving mechanism and can be driven to perform linear motion along the X/Y/Z directions while the grooving cutter is driven to rotate at a high speed; the X/Y linkage is used for processing and obtaining a continuous track of a crack center, the Z direction is used for controlling the depth of a groove, the bottom camera is used for shooting a crack image of a vehicle bottom area in a vehicle parking state, and the bottom light source provides illumination for the bottom camera.
Example 2:
in order to realize the automatic navigation of the grooving robot trolley, two working modes are respectively designed, one is a continuous grooving working mode, as shown in fig. 3-1; the other is the intermittent slotting mode, as shown in fig. 3-2.
(1) The continuous mode of operation is the most intuitive and common mode of operation. The mode simulates manual slotting equipment, and the slotting robot trolley performs slotting operation while performing line inspection navigation; line patrol navigation is realized by a head camera; the running track of the slotting cutter is the center track of the crack. The slotting cutter does not require a cutter position adjustment mechanism in addition to the rotary primary motion.
The working mode has the advantages that the robot trolley is simple in structure and high in slotting efficiency; the disadvantage is that it is difficult to meet the grooving accuracy requirement for the following reasons:
the main factors influencing slotting accuracy are line inspection errors, which are mainly influenced by two parts, namely an imaging and center extraction algorithm error delta 1 of a navigation camera, which is generally smaller; secondly, the walking error delta 2 of the robot trolley is affected by factors such as the weight, inertia, ground friction, vibration and the like of the slotting trolley, the walking error delta 2 is far greater than delta 1, and in general, the error delta 2 only can far exceed the slotting precision requirement. And the actual slotting accuracy error is the sum of the two parts. Therefore, the working mode is difficult to meet the grooving precision requirement.
(2) The intermittent slotting working mode of the application is shown in fig. 3-2. The robot trolley is navigated by the head navigation camera, after the robot trolley advances by one vehicle body length, the trolley is stopped, the bottom camera shoots, a crack image covered by the bottom of the trolley is obtained, the center of a track is extracted, and the center track provides a feeding instruction in the X/Y/Z direction for a slotting cutter driving mechanism, so that slotting is completed along the center of the crack by the slotting cutter.
The working mode has the defects that the structure of the robot trolley is relatively complex, and the slotting efficiency is relatively low; the grooving device has the advantages of high grooving precision, and the reason is as follows:
under the working mode, the grooving operation is performed after the line inspection is performed, and the trolley is stopped, so that the cruise camera error delta 1 and the robot walking error delta 2 do not influence the grooving precision. The slotting precision is affected by two parts, namely, the error delta 3 of bottom camera imaging and crack center detection is small; secondly, under the existing technical conditions, even a low-cost three-dimensional movement mechanism can achieve high movement precision, so that the error delta 4 is small. Therefore, the intermittent working mode designed by the application has the grooving precision far higher than that of the continuous working mode, and can meet the grooving requirement of highway cracks, so that the intermittent working mode is adopted by the application.
Aiming at the grooving working mode and the robot structural design, the working method of the automatic grooving robot for the highway cracks comprises the following steps:
(1) Guiding the slotting robot trolley to a crack starting position (such as the lower position of the figure 3-2) so that a crack is positioned between wheels on two sides and the crack is positioned in the field of view of the navigation camera;
(2) The robot trolley starts a slotting mode;
(3) A navigation camera at the front part of the trolley starts to collect a road surface image at the front part of the trolley, extracts a center line of a crack and obtains a center track image of the crack;
(4) The robot automatically detects the deviation between the current pose and the center position of the crack, automatically tracks the center of the crack, and stops after controlling the trolley to move forward by one vehicle body distance;
(5) Starting a bottom camera by the robot trolley, shooting a bottom crack image, extracting a crack center line, and obtaining a crack center track image of a vehicle bottom area;
(6) The robot controls the slotting cutter driving mechanism to drive the slotting cutter to move, slotting is completed along the center track of the crack obtained in the step (5), and the slotting cutter position is reset (the initial position in the X/Y/Z three-dimensional direction);
(7) And (3) judging whether a crack center line exists in the visual field of the navigation camera, if so, returning to the step (3), continuing the slotting work of the next parking space, and realizing autonomous navigation through a crack center line extraction method and a crack center automatic tracking method. The method comprises the steps of carrying out a first treatment on the surface of the If the navigation camera has no crack center line in the field of view, the robot stops working and sends task completion alarm information.
In the step (3), the crack center line extraction method comprises the following steps:
(3-1) reading in the road crack image acquired by the navigation camera, and converting the road crack image into a gray image, specifically:
f(u,v)＝0.2989*r(u,v)+0.587*g(u,v)+0.114*b(u,v) (1)
wherein f (u, v) is a gray scale image matrix; r (u, v), G (u, v), B (u, v) are R/G/B component matrices of RGB images, respectively, the result is shown in fig. 4;
(3-2) gaussian filtering the slit gray image:
F(u,v)＝G(u,v)*f(u,v) (2)
wherein ,the method is characterized in that the method is a Gaussian filter kernel function, u is the number of lines of an image, v is the number of columns of the image, and sigma is the standard deviation of the kernel function; the filter size is 7x7, the standard deviation is σ=0.5-3.5, in this embodiment, the standard deviation is σ=2; the results are shown in FIG. 5;
(3-3) performing threshold segmentation on the filtered image to obtain a binary image, wherein the specific method comprises the following steps of:
wherein t=255×0.1=25.5; the results are shown in FIG. 6;
(3-4) carrying out morphological processing on the binary image, removing noise points, and obtaining a crack center image, wherein the specific process is as follows:
1) The 8 connected domain marking is carried out on the binary image h (u, v), specifically:
let (u, v) be the center pixel, search and mark the connected domain for 8 neighborhood including (u+1, v), (u-1, v), (u, v+1), (u, v-1), (u+1, v+1), (u-1, v+1) and (u-1, v-1), the specific method is:
a. progressive scanning image, comprising a sequence of successive white pixels in each row, each sequence being assigned a reference j i ，i＝1,2，3...；
b. Judging whether the sequence of the line is communicated with the sequence in the previous line one by one from the second line of the image, namely judging whether each white pixel point in the sequence has a white point in the 8 neighborhood of the previous line, if not, representing non-communication, and if so, representing communication; if not, it is given a new reference (the largest reference in the previous row is j i Then the new reference sign is j i+1 ) The method comprises the steps of carrying out a first treatment on the surface of the If it is connected to only 1 sequence in the previous row, the sequence is combined with the sequence of the previous row to form a new sequence, and the new sequence is given the reference number of the sequence of the previous row (the previous row has two sequences in total and the reference numbers are j respectively i and ji+1 If a certain sequence of this row is numbered j in the previous row i The sequences connected in the two rows are combined and labeled j i Assign a new sequence); if the sequence is communicated with 2 or more sequences in the previous row, combining the sequence with all the communicated sequences to form a new sequence, and assigning the smallest number in the sequence in the previous row to the new sequence (the two sequences in the previous row are respectively numbered j i and ji+1 If a certain sequence in the row is identical to the two sequences in the previous rowAll connected sequences in both rows are combined and the smallest number, j i Assigned to a new sequence) and writing the labels of the sequences already connected in the previous row to the equivalent pair < j i ,j i+1 ,. >, indicating that the sequences marked by these numbers belong to the same sequence;
c. and judging all equivalent pairs, and if the two equivalent pairs have the same reference numerals, merging the two equivalent pairs into a new equivalent pair because the sequences marked by the two equivalent pairs belong to the same connected domain. Finally, each new equivalent pair is marked with a connected domain label, denoted as L i The method comprises the steps of carrying out a first treatment on the surface of the At the same time, the number of connected domain pixels marked by the label of each connected domain is recorded and is recorded as N i The method comprises the steps of carrying out a first treatment on the surface of the Wherein i=1..m is the number of connected domains in the image; then n= [ N ] 1 N 2 ...N m ]Representing the number of pixel points of each connected domain in the image; with L= [ L ] 1 L 2 ...L m ]A reference numeral representing a corresponding one of the connected domains;
2) According to the number N of the pixel points of each connected domain i The communicating domains are ordered in descending order, [ N ', L ]']＝Rank(N,'descend')
Wherein, the number N of each connected domain pixel in N' is i Descending order from big to small; the element in L' is each N i The labels of the corresponding connected domains; 'desend' represents a descending order;
3) The first 10 connected domains N with the largest pixel number are reserved 1 ～N 10 I.e. in N 10 For the segmentation threshold, when the number of the pixels of the connected domain is larger than the segmentation threshold, reserving the connected domain; if the number of the pixels of the communication area is smaller than the segmentation threshold value, discarding the communication area; after the self-adaptive threshold segmentation step, a binary image H (u, v) of the crack is obtained as shown in fig. 7;
4) Performing closed operation to connect broken cracks; the specific operation steps are as follows:
the image H (u, v) is inflated by a disk structure S with a radius of 5 pixels, and then etched,
wherein ,Θ is the image expansion and corrosion symbol, respectively; the results are shown in FIG. 8.
5) Performing skeleton extraction on the image obtained in the step 4), and extracting a central pixel point of a crack, wherein the method comprises the following specific steps of:
assuming that the image matrix obtained through the above closed operation is H (u, v), the crack center image can be obtained by the following method:
wherein ,is a subset of the fracture center image, which can be obtained by combining the subset. Wherein Θ, & gt>Respectively image erosion and open operation symbols; (H Θkb) represents that the image H (u, v) is subjected to successive k times of etching with the structure B, which is a circular structure having a radius of 7 pixels; then, performing open operation on the corroded image, eliminating a small white area, and solving the difference between the corroded image and the corroded image, so as to obtain a crack center subset image; preferably, the number of times of k is determined by: is the last iteration before H (u, v) is eroded to empty set, i.eThen, the image S (H) processed as described above represents a skeleton image of the image H (u, v), denoted as S (u, v), and represents an image matrix of the center of the crack, as shown in fig. 9.
In the step (4), the automatic tracking method of the crack center comprises the following steps:
as shown in FIG. 10, a device is provided withFixed O W -X W Y W Z W Is the world coordinate system, O V -X V Y V Z V Is the car coordinate system, Z W and ZV The axes being respectively with O W -X W Y W and OV -X V Y V The plane is vertical and is not shown in fig. 10; o-uv is the image coordinate system; the image S (u, v) relates to X V Axisymmetric, and the image closest to the row and X of the trolley V The intersection point of the axes is denoted as X;
(4-1) in the broken line box of fig. 10, a centerline image S (u, v) is extracted by the method for extracting a centerline of a crack in step (3) of the design of the present application, any point on the image being defined as S (u, v);
(4-2) at any time, taking the intersection point S of the crack center and the image center line 1 (u 1 ,v 1 ) For the current navigation tracking point, the corresponding course angle of the point is phi, the point is defined as a tracking starting point, and the center line of the image is parallel to the v axis along the u axis, as shown by a broken line in fig. 10;
(4-3) to achieve crack center tracking, it is required to control the speed V of the cart X Along X V By controlling the angular velocity omega of the heading while the shaft is advancing Z The tracking start point is made to fall at the center S 'of the image S (u, v)' 1 (u 1 ,v 1 ) The point is defined as a navigation tracking target point;
(4-4) in actual control, since the forward speed V of the carriage X Is known and adjustable about Z V Heading angular velocity ω of shaft Z Is determined by the following method:
let it be assumed that the start point S is tracked 1 (u 1 ,v 1 ) There is a mapping point S 0 (u 0 ,v 0 ) Wherein |XS 0 |＝|XS 1 I, the distance from the mapping point to the tracking start point to the X point is the same; assuming that the tracking process is completed within Δt, this means that the cart follows X within Δt V The axis is moved by a distance |S '' 1 S 0 Simultaneous with I along Z V The heading angle is phi by rotating the shaft, and the heading angular velocity can be calculated by the following formula:
wherein, |XS' 1 The actual length of the I in the world coordinate system can be obtained through calibration of an imaging system;
(4-5) controlling start and stop of the robot trolley according to the following method: as long as the crack centerline trajectory can be detected in the crack centerline image, the robot trolley always determines V according to the methods of steps (4-1) - (4-4) X and ωZ Advancing; when no crack central line track exists in the visual field image shot by the trolley navigation camera, V is determined X and ωZ All set to 0 and the robot trolley is stopped.
Example 3:
the utility model provides an automatic slotting robot of highway crack is in application of integration highway repair system, integration highway repair system includes automatic slotting robot of highway crack, crack filling device, and automatic slotting robot of highway crack is used for carrying out automatic slotting to the highway crack, and crack filling device is used for the crack filling after the slotting.

Claims (4)

1. A working method of a highway crack automatic slotting robot is characterized in that,
the automatic slotting robot for the highway cracks comprises a robot trolley, wherein a chassis of the robot trolley comprises two rear wheels driven by differential speed and two front wheels for controlling steering;
the front part of the robot trolley is provided with a navigation camera, a head light source and a laser radar, a crack image acquired by the navigation camera is extracted from the center and then used for navigation of the robot, the head light source is used for navigation illumination, and the laser radar is used for automatic obstacle avoidance of the robot trolley;
a slotting cutter, a bottom camera and a bottom light source are arranged below the chassis of the robot trolley, and the slotting cutter is a tooth saw or a milling cutter; the grooving cutter is driven by the grooving cutter driving mechanism and is used for driving the grooving cutter to realize high-speed rotation and simultaneously driving the grooving cutter to perform linear motion along the X/Y/Z directions; the X/Y linkage is used for processing and obtaining a continuous track of a crack center, the Z direction is used for controlling the depth of a slot, the bottom camera is used for shooting a crack image of a bottom area of the trolley in a trolley parking state, and the bottom light source is used for providing illumination for the bottom camera;
the working method comprises the following steps:
(1) Guiding the slotting robot trolley to a crack starting position, so that a crack is positioned between wheels on two sides and is positioned in a field of view of a navigation camera;
(2) The robot trolley starts a slotting mode;
(3) A navigation camera at the front part of the trolley starts to collect a road surface image at the front part of the trolley, extracts a center line of a crack and obtains a center track image of the crack;
(4) The robot automatically detects the deviation between the current pose and the center position of the crack, automatically tracks the center of the crack, and stops after controlling the trolley to move forward by one vehicle body distance;
the automatic tracking method for the crack center comprises the following steps:
setting O W -X W Y W Z W Is the world coordinate system, O V -X V Y V Z V Is the car coordinate system, Z W Shaft and O W -X W Y W The plane is vertical, Z V Shaft and O V -X V Y V The plane is vertical; o-uv is the image coordinate system; the image S (u, v) relates to X V Axisymmetric, and the image closest to the row and X of the trolley V The intersection point of the axes is denoted as X;
(4-1) extracting a center line image S (u, v) by using the crack center line extraction method in the step (3), wherein any point on the image is defined as S (u, v);
(4-2) at any time, taking the intersection point S of the crack center and the image center line 1 (u 1 ,v 1 ) The method comprises the steps that as a current navigation tracking point, a heading angle corresponding to the point is phi, the point is defined as a tracking starting point, and an image center line is parallel to a v axis along a u axis;
(4-3) to achieve crack center tracking, it is required to control the speed V of the cart X Along X V By controlling the angular velocity omega of the heading while the shaft is advancing Z The tracking start point is made to fall at the center S of the image S (u, v) 1 ’(u 1 ,v 1 ) The point is defined as a navigation tracking target point;
(4-4) in actual control, the forward speed V of the carriage X Is known and adjustable about Z V Heading angular velocity ω of shaft Z Is determined by the following method:
tracking the starting point S 1 (u 1 ,v 1 ) There is a mapping point S 0 (u 0 ,v 0 ) Wherein |XS 0 |＝|XS 1 I, the distance from the mapping point to the tracking start point to the X point is the same; completing the tracking process within Δt means that the trolley follows X within Δt V Distance of axial linear movement |S 1 'S 0 Simultaneous with I along Z V The shaft rotates the course angle phi, and the course angular velocity is calculated by the following formula:
wherein ,|XS1 The actual length of' |in the world coordinate system is obtained through calibration of an imaging system;
(4-5) controlling start and stop of the robot trolley according to the following method: as long as the crack centerline trajectory can be detected in the crack centerline image, the robot trolley always determines V according to the methods of steps (4-1) - (4-4) X and ωZ Advancing; when no crack central line track exists in the visual field image shot by the trolley navigation camera, V is determined X and ωZ All set to 0, the robot trolley stops;
(5) Starting a bottom camera by the robot trolley, shooting a bottom crack image, extracting a crack center line, and obtaining a crack center track image of a vehicle bottom area;
(6) The robot controls the slotting cutter driving mechanism to drive the slotting cutter to move, slotting is completed along the center track of the crack obtained in the step (5), and the slotting cutter is reset;
(7) Judging whether a crack center line exists in the visual field of the navigation camera, if yes, returning to the step (3), continuing the slotting work of the next parking space, and realizing autonomous navigation through a crack center line extraction method and a crack center automatic tracking method; if the navigation camera has no crack center line in the field of view, the robot stops working and sends task completion alarm information.
2. The method for operating a highway crack automatic slotting robot according to claim 1, wherein in the step (3), the crack center line extraction method comprises the steps of:
(3-1) reading in the road crack image acquired by the navigation camera, and converting the road crack image into a gray image, specifically:
f(u,v)＝0.2989*r(u,v)+0.587*g(u,v)+0.114*b(u,v) (1)
wherein f (u, v) is a gray scale image matrix; r (u, v), G (u, v), B (u, v) are R/G/B component matrices of RGB images, respectively;
(3-2) gaussian filtering the slit gray image:
F(u,v)＝G(u,v)*f(u,v) (2)
wherein ,the method is characterized in that the method is a Gaussian filter kernel function, u is the number of lines of an image, v is the number of columns of the image, and sigma is the standard deviation of the kernel function; the filter size is 7x7, and the standard deviation is sigma=0.5-3.5;
(3-3) performing threshold segmentation on the filtered image to obtain a binary image, wherein the specific method comprises the following steps of:
wherein t=255×0.1=25.5;
(3-4) carrying out morphological processing on the binary image, removing noise points, and obtaining a crack center image, wherein the specific process is as follows:
1) The 8 connected domain marking is carried out on the binary image h (u, v), specifically:
let (u, v) be the center pixel, search and mark the connected domain for 8 neighborhood including (u+1, v), (u-1, v), (u, v+1), (u, v-1), (u+1, v+1), (u-1, v+1) and (u-1, v-1), the specific method is:
a. progressive scanning image, comprising a sequence of successive white pixels in each row, each sequence being assigned a reference j i ，i＝1,2，3...；
b. Judging whether the sequence of the line is communicated with the sequence in the previous line one by one from the second line of the image, namely judging whether each white pixel point in the sequence has a white point in the 8 neighborhood of the previous line, if not, representing non-communication, and if so, representing communication; if not, a new label is given to it; if it is connected with only 1 sequence in the last row, combining the sequence with the sequence of the last row to form a new sequence, and assigning the label of the sequence of the last row to the new sequence; if the sequence is communicated with 2 or more sequences in the previous row, combining the sequence with all communicated sequences to form a new sequence, assigning the smallest number in the sequence in the previous row to the new sequence, and writing the number of the sequence which is communicated in the previous row into an equivalent pair < j i ,j i+1 ,. >, indicating that the sequences marked by these numbers belong to the same sequence;
c. judging all equivalent pairs, if the two equivalent pairs have the same reference numerals, combining the two equivalent pairs into a new equivalent pair, and finally marking each new equivalent pair with a connected domain reference numeral L i The method comprises the steps of carrying out a first treatment on the surface of the At the same time, the number of connected domain pixels marked by the label of each connected domain is recorded and is recorded as N i The method comprises the steps of carrying out a first treatment on the surface of the Wherein i=1..m is the number of connected domains in the image; then use n= [ N ] 1 N 2 ...N m ]Representing the number of pixel points of each connected domain in the image; with L= [ L ] 1 L 2 ...L m ]A reference numeral representing a corresponding one of the connected domains;
2) According to the number N of the pixel points of each connected domain i The communicating domains are ordered in descending order, [ N ', L ]']＝Rank(N,'descend')
Wherein, the number N of each connected domain pixel in N' is i Descending order from big to small; the element in L' is each N i The labels of the corresponding connected domains; 'desend' tableShowing a descending order of arrangement;
3) The first 10 connected domains N with the largest pixel number are reserved 1 ～N 10 I.e. in N 10 For the segmentation threshold, when the number of the pixels of the connected domain is larger than the segmentation threshold, reserving the connected domain; if the number of the pixels of the communication area is smaller than the segmentation threshold value, discarding the communication area; obtaining a binary image H (u, v) of the crack after the self-adaptive threshold segmentation step;
4) Performing closed operation to connect broken cracks; the specific operation steps are as follows:
the image H (u, v) is inflated by a disk structure S with a radius of 5 pixels, and then etched,
wherein ,the symbol is an image expansion symbol, and Θ is an image corrosion symbol;
5) Performing skeleton extraction on the image obtained in the step 4), and extracting a central pixel point of a crack, wherein the method comprises the following specific steps of:
assuming that the image matrix obtained through the above closed operation is H (u, v), the crack center image can be obtained by the following method:
wherein ,is a subset of the crack center image, and the crack center image is obtained through the subset and operation; wherein Θ is an image corrosion symbol, < +.>Is an open operation symbol; (H. Theta. KB) represents the use of knotsThe structure B is a circular structure with a radius of 7 pixels, and the structure B is used for corroding the image H (u, v) for k times continuously; then, performing open operation on the corroded image, eliminating a small white area, and solving the difference between the corroded image and the corroded image, so as to obtain a crack center subset image; the number of times of k is determined by: is the last iteration before H (u, v) is eroded to empty set, i.e
Then, the image S (H) processed as described above represents a skeleton image of the image H (u, v), denoted as S (u, v), and represents an image matrix of the center of the crack.
3. The method of claim 2, wherein in the step (3-2), the standard deviation is σ=2.
4. An integrated highway repair system, the integrated highway repair system comprises a highway crack automatic slotting robot and a crack filling device, wherein the highway crack automatic slotting robot is used for automatically slotting a highway crack by using the working method of claim 1, and the crack filling device is used for filling the slotted crack.
CN202210974154.5A 2022-08-15 2022-08-15 Automatic slotting robot for highway cracks and working method and application thereof Active CN115294080B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210974154.5A CN115294080B (en) 2022-08-15 2022-08-15 Automatic slotting robot for highway cracks and working method and application thereof

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210974154.5A CN115294080B (en) 2022-08-15 2022-08-15 Automatic slotting robot for highway cracks and working method and application thereof

Publications (2)

Publication Number Publication Date
CN115294080A CN115294080A (en) 2022-11-04
CN115294080B true CN115294080B (en) 2023-09-08

Family

ID=83829663

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210974154.5A Active CN115294080B (en) 2022-08-15 2022-08-15 Automatic slotting robot for highway cracks and working method and application thereof

Country Status (1)

Country Link
CN (1) CN115294080B (en)

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105975972A (en) * 2016-04-27 2016-09-28 湖南桥康智能科技有限公司 Bridge crack detection and characteristic extraction method based on image
CN106514607A (en) * 2016-12-30 2017-03-22 成都圭目机器人有限公司 Four-wheel drive road face repairing robot system and road face repairing method
EP3454159A1 (en) * 2017-09-06 2019-03-13 Alerion Technologies, S.L. Method and device of autonomous navigation
CN110084302A (en) * 2019-04-26 2019-08-02 西安因诺航空科技有限公司 A kind of crack detection method based on remote sensing images
CN111223354A (en) * 2019-12-31 2020-06-02 塔普翊海（上海）智能科技有限公司 Unmanned trolley, and AR and AI technology-based unmanned trolley practical training platform and method
CN111287064A (en) * 2020-03-24 2020-06-16 长安大学 Pavement crack grooving machine and grooving method
CN111310558A (en) * 2019-12-28 2020-06-19 北京工业大学 Pavement disease intelligent extraction method based on deep learning and image processing method
CN112419250A (en) * 2020-11-13 2021-02-26 湖北工业大学 Pavement crack digital image extraction, crack repair and crack parameter calculation method
CN113358659A (en) * 2021-04-25 2021-09-07 上海工程技术大学 Camera array type imaging method for automatic detection of high-speed rail box girder crack
CN114072698A (en) * 2019-06-28 2022-02-18 伟摩有限责任公司 Beam homogenization for anti-occlusion

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101389244B1 (en) * 2013-04-18 2014-04-24 재단법인대구경북과학기술원 Robot for managing of structure and method controling thereof

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105975972A (en) * 2016-04-27 2016-09-28 湖南桥康智能科技有限公司 Bridge crack detection and characteristic extraction method based on image
CN106514607A (en) * 2016-12-30 2017-03-22 成都圭目机器人有限公司 Four-wheel drive road face repairing robot system and road face repairing method
EP3454159A1 (en) * 2017-09-06 2019-03-13 Alerion Technologies, S.L. Method and device of autonomous navigation
CN110084302A (en) * 2019-04-26 2019-08-02 西安因诺航空科技有限公司 A kind of crack detection method based on remote sensing images
CN114072698A (en) * 2019-06-28 2022-02-18 伟摩有限责任公司 Beam homogenization for anti-occlusion
CN111310558A (en) * 2019-12-28 2020-06-19 北京工业大学 Pavement disease intelligent extraction method based on deep learning and image processing method
CN111223354A (en) * 2019-12-31 2020-06-02 塔普翊海（上海）智能科技有限公司 Unmanned trolley, and AR and AI technology-based unmanned trolley practical training platform and method
CN111287064A (en) * 2020-03-24 2020-06-16 长安大学 Pavement crack grooving machine and grooving method
CN112419250A (en) * 2020-11-13 2021-02-26 湖北工业大学 Pavement crack digital image extraction, crack repair and crack parameter calculation method
CN113358659A (en) * 2021-04-25 2021-09-07 上海工程技术大学 Camera array type imaging method for automatic detection of high-speed rail box girder crack

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
基于神经网络深度学习的无人驾驶小车设计与制作;徐中林等;《电子制作》(第22期);37-40 *

Also Published As

Publication number Publication date
CN115294080A (en) 2022-11-04

Similar Documents

Publication Publication Date Title
CN112498367B (en) Driving track planning method and device, automobile, controller and computer readable storage medium
CN111071249B (en) Unmanned vehicle emergency steering obstacle avoidance method based on obstacle motion prediction
CN111736603A (en) Unmanned sweeper and long-distance welting sweeping method thereof
CN112319429B (en) Control system and control method for automatically cleaning front windshield of bullet train
CN111967360A (en) Target vehicle attitude detection method based on wheels
CN211729203U (en) Runway laser degumming robot
CN211734978U (en) Unmanned rapid comprehensive road detection vehicle system
CN110842918B (en) Robot mobile processing autonomous locating method based on point cloud servo
CN111230903A (en) Runway laser degumming robot
FR3112502B1 (en) Automatic control of a part deburring system
CN111123950A (en) Driving control method and device and vehicle
CN110262504A (en) A kind of adjustable multilasered optical radar coupled system of structure and its control method
CN115294080B (en) Automatic slotting robot for highway cracks and working method and application thereof
CN112819766A (en) Bridge defect overhauling method, device, system and storage medium
CN111776243A (en) Control method and device for boarding bridge
CN110716563A (en) Electronic map given trajectory-based electric wheelchair path tracking control method and device
CN111287064A (en) Pavement crack grooving machine and grooving method
CN109703553B (en) Automatic parking method based on traction point tracking
CN114646976A (en) Single-line laser radar road surface detection device based on holder and detection method thereof
WO2022252220A1 (en) Precise stopping system and method for multi-axis flatbed vehicle
CN113341999A (en) Forklift path planning method and device based on optimized D-x algorithm
CN112620253A (en) Rust removal equipment
CN115328135B (en) Scribing robot calibration and scribing method and device, electronic equipment and scribing robot
CN115285110A (en) Parking path planning method and system based on sine curve
CN201735929U (en) Steel plate outline scanning mechanism used on steel plate cutting machine

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant